The Arkansas Payment Improvement Initiative: Early Perceptions of Multi-Payer Reform in a Fragmented Provider Landscape

June 12, 2015
Michael E. Chernew, PhD

,
William E. Golden, MD

,
Christopher H. Mathis, JD, MPA

,
A. Mark Fendrick, MD

,
Michael W. Motley, MPH

,
Joseph W. Thompson, MD, MPH

The American Journal of Accountable Care, June 2015, Volume 3, Issue 2

Arkansas has implemented multi-payer payment reform incorporating both episodic and Patient-Centered Medical Home models. Early perceptions of a sample of stakeholders were largely positive to date.

As efforts to move away from fee-for-service payment have accelerated on both federal and state levels, public and private payers have been experimenting with different payment models. There is tremendous heterogeneity in approaches across states1-3; for example, some of these reforms only focus on Medicaid, while some include additional payers. Some rely on global payments and some on bundled payments for selected episodes.4-11

Arkansas’ approach is novel in several ways. The Arkansas Payment Improvement Initiative (APII) is a multi-payer model combining patient-centered medical homes (PCMHs) with episodic payments for certain acute and chronic conditions.12 Provider participation in the PCMH is voluntary, and the episode payments system is mandatory for most populations and all providers. A health home component for individuals with more complex care needs is planned but not yet implemented.

Viewing the APII through a national lens, the Arkansas model could turn out to be a particularly useful bellwether for other states, especially less populated states with a more fragmented and largely rural provider environment.

Arkansas’ reforms also reflect cooperation between and among providers, insurers, employers, and state government officials. Regular planning and informational meetings in addition to outreach with stakeholders, as well as consensus on a general framework for reform, have been hallmarks of the program’s development.

This paper reports on early perceptions of how the model is working. We interviewed a convenience sample of Arkansas stakeholders; provider perceptions are based on interviews with representatives of the state medical society and hospital association, who have been canvassing their members to get a broad perception of provider reaction, and providers in 3 Arkansas practices, both urban and rural. We also interviewed representatives from 2 private insurers that collectively cover almost 90% of the commercially insured population, as well as large employers in the state. Our interviews were unstructured; they focused on reaction to the APII, whether or not providers are changing practice as a result of the new approach, and if so, how and for whom? Because health homes have not yet been implemented, we focus here on episodes and medical homes.

The APII

Episodes of Care

Arkansas has implemented episodic payment for certain conditions.13 The initial 5 episodes, launched at the end of 2012, were: upper respiratory infections (URIs), attention deficit/hyperactivity disorder (ADHD), hip and knee replacement, perinatal (pregnancy), and congestive heart failure (CHF). Three more—colonoscopy, cholecystectomy (gallbladder removal), and tonsillectomy—were designed and implemented in 2013. An additional 2—coronary artery bypass grafting (CABG) and asthma—were introduced in 2014, and 5 more—including oppositional defiant disorder (ODD), percutaneous coronary intervention, chronic obstructive pulmonary disease, and ADHD/ODD comorbidity—will likely launch in 2015.

For each episode of care, a Principal Accountable Provider (PAP) is assigned by payers based on analysis of claims data. The PAP is typically the primary decision maker for most of the care delivered within an episode and often has the greatest ability to influence and improve patient care. For example, in the perinatal episode, the PAP is the delivering physician. Only 1 PAP is assigned for each episode and held accountable for all care delivered within that episode over a predetermined time period. Most, but not all, episodes involve acute conditions that have clear triggering events, such as birth or surgery. In the case of a chronic illness, such as ADHD, the PAP is retrospectively assigned based on provision of the majority of services over the episode’s duration.

PAPs may be rewarded, penalized, or remain financially neutral based on how the average costs for their episodes compare with thresholds predetermined by payers. Specifically, for each episode, there are thresholds for “acceptable,” “commendable,” and “gain sharing limit” expenditures. These thresholds are determined separately by each payer in order to adhere to anti-trust regulations, and they are communicated to providers prior to the start of an episode performance period based on a retrospective review of Arkansas expenditures for each episode over a predetermined performance period (generally 12 months). Services within the episode for all providers are reimbursed through each payer’s fee-for-service fee schedule.

At the conclusion of a performance period, an episode profile, which includes all costs and quality targets, is calculated for each PAP. PAPs with average costs above the acceptable threshold for an episode are assessed a penalty payment. PAPs with average costs between the acceptable and commendable thresholds neither pay a penalty nor receive gain sharing. PAPs with average costs below the commendable threshold are eligible for gain sharing—those with average costs below the gain sharing limit are still eligible for gain sharing if they meet quality metrics, but derive no additional financial benefit for savings beyond the limit.

Guidelines for each episode also require that certain quality measures be passed in order to qualify for gain sharing, and that additional associated quality measures be tracked and reported to PAPs. Operationally, all providers involved in the patient’s care (including the PAP) continue to be paid on a fee-for-service basis during the episodes, but it is the practice employing the PAP that settles up at the end of the performance period and bears the risk of high total spending (or shares the gain of lower spending).

Patient-Centered Medical Homes

In addition to these episodic payment reforms, all Arkansas practices that serve as primary care provider to at least 300 Medicaid patients are eligible to participate as a PCMH as of January 1, 2014.14 Participating practices receive a per member per month (PMPM) medical home support payment to facilitate care coordination and practice transformation.

In order to receive these monthly PMPMs, practices must demonstrate that they have implemented and are performing numerous activities integral to building a medical home structure. These activities include providing 24/7 live voice access to a health professional, identification of and formulation of care plans for high-risk patients, flexible same-day scheduling, installment of meaningful use certified electronic health records, assessment of operations and opportunities for improvement, and other practice enhancements related to a PCMH framework.

An average $4 PMPM payment ($7 for adults and $3 for children) is paid by Arkansas Medicaid directly to practices to be used for new investments in clinical operations that achieve effective care coordination, patient engagement, and improved outcomes. As one example, a practice with a Medicaid-eligible panel of 3000 patients could receive an average of $144,000 in additional prospective practice support payments from Medicaid over a 12-month period.

To supplement these ongoing care coordination payments, Medicaid pays an additional $1 PMPM payment directly to a technical support vendor on behalf of practices that have chosen to participate in supplemental training. These payments are for a limited duration of 24 months and are designed to catalyze PCMH transformation. Technical support includes 30 hours of individual coaching, drafting of a PCMH needs assessment and implementation plan, access to online training, and peer-to-peer networking sessions. In 2014, the state certified Qualis Health as a PCMH technical support vendor. A second vendor, the Arkansas Foundation for Medical Care (AFMC), was added in 2015.

The PCMH program offers practices an opportunity to receive shared savings (adjusted to avoid double counting with the episode payment shared savings) if certain requirements are met. Specifically, the PCMH model is upside-only and practices are eligible to retain up to 50% of savings by spending below either a preset statewide threshold or a threshold based on improvement over their own performance in previous years. To be eligible to receive shared savings, a practice must have 5000 Medicaid patients who have been attributed for at least 6 months. As many practice sites lack a sufficient number of patients and doctors to reach this threshold, Arkansas Medicaid initially permitted 2 practices or multiple providers with the same tax identification number to pool their patients to achieve the 5000-patient minimum enrollment. These pooled practices must comply with combined quality metrics and possess the required attributes of a medical home. Starting in 2015, more than 2 practices will be allowed to pool and there will also be a statewide default pool, which will provide another means for practice eligibility for shared savings under the Medicaid program. A shared savings component involving participating private payers is scheduled to begin in 2016.

Who Is Participating?

Arkansas’ approach to both the APII’s episodic payment and PCMH reforms has achieved participatory support from multiple payers. For episodes, in addition to Medicaid, QualChoice of Arkansas and Arkansas Blue Cross and Blue Shield (the state’s largest commercial insurer with over 80% of the private market) are participating. While participating payers have agreed on the design of each episode, the private insurers do not participate in all episodes. For example, neither Blue Cross nor QualChoice is using episodic payment for URI or ADHD, though Medicaid does, and Blue Cross and Medicaid use episodic payment for CHF, tonsillectomy, coronary artery bypass graft surgery, and asthma, while QualChoice does not. Another of the top 3 commercial payers, UnitedHealthcare, is not yet directly involved in episodes.

Among self-insured employers, the largest private employer, Walmart, and the 2 largest public plans, the State Employee Plan and Public School Employee Plan, actively participate in episodic payment. In 2015, Blue Cross Blue Shield extended episodes to their fully insured products and all self-insured contracts. In all, roughly three-fourths of commercially covered Arkansas lives are subject to the episodic payment component of the APII.

With respect to the PCMH component, multiple payers initiated participation in the Comprehensive Primary Care (CPC) Initiative medical home program in 2013, including QualChoice, Blue Cross Blue Shield, Humana, Medicaid, and Medicare. Self-insured payers Blue Cross Blue Shield (for its own employee plan) and Walmart, as well as the State Employee Plan and Public School Employee Plan, joined CPC through participation with PMPM care coordination fees for their covered populations in 2014. Medicaid has expanded this primary care participation by establishing its own PCMH model, building on CPC with extended requirements for transformational milestones and incorporating its shared savings strategy through a legislatively approved rule. Initiated in 2014, PCMH participation has exceeded primary care enrollment targets.

In December 2014, the Arkansas Insurance Department established rules for qualified health plans (QHPs) participating in the health insurance marketplace that require PCMH participation in 2015. This requirement is a component of the Arkansas Health Care Independence Program, the Medicaid expansion premium assistance plan also known as the “private option.” As a result, QualChoice, Blue Cross Blue Shield, and Ambetter (Centene) will attribute all of their roughly 250,000 exchange enrollees and support PCMH transformation through an average $5 PMPM payment. These payers will not pay the additional PMPM for practice transformation and technical support but will use quality assurance assessments conducted by Medicaid for deemed provider eligibility. In addition to this new QHP participation, Blue Cross is extending PCMH participation across its entire fully insured book of business in 2015.

Results

Provider Reaction

The multi-payer approach has generated a certain sense of inevitability that comes from most payers adopting a consistent payment model. Physicians wary of reform altogether, or at least of the permanence of the new model, may perceive they may have little choice but to adapt to mandatory episodes and strongly consider PCMH participation. With respect to episodes, one provider noted, “Doctors are just starting to realize that it’s not going away and [are] wrapping their minds around it.” An interviewee from the hospital sector noted that embracing the APII made a virtue of necessity in that it was the most palatable of the 3 options under consideration to deal with a significant fiscal shortfall in the Medicaid program, the other 2 being deep across-the-board fee cuts or potentially even deeper cuts if an out-of-state managed care company were brought in to oversee the program. As this individual put it, “The other alternatives were much worse.”

A notable factor mentioned with respect to provider reaction and receptivity to change is that Arkansas Medicaid is “easy to work with…they reimburse faster than private payers and have made it easy to be a physician within the program.” By contrast, some physicians and hospitals had been leery of having private payers participate due to a concern that they would take the opportunity to lower their rates to be more in line with Medicaid rates and not just adopt the same payment methodology. Yet, those fears have faded now that the program is in operation.

While the individual components of the APII are well delineated, there is still some confusion about the interaction of episodes and medical homes. Also, some skepticism exists with respect to the appropriateness of the episodic model under certain conditions. One of our interviewees, though generally supportive of episodes, noted that a chronic condition such as ADHD is “not an episode” and would more appropriately be dealt with within the context of a PCMH.

The financial impact on practices differs with respect to episodes and medical homes. As one individual we interviewed stated in noting the difference, “Episodes bite whereas PCMH doesn’t.” One physician told us, “Care coordination and practice transformation take a lot of time and energy and it’s nice to have the money to finally do it.” At the same time, there is only minimal added cost to the practice to deal with episodes, whereas PCMH involves an investment in transformation. In the view of one individual we interviewed, “A practice generally needs to generate shared savings [in addition to support payments] or you have lost money.”

Our interviews suggest some of the most facile transformation to the PCMH model has often come from smaller practices, which are more nimble. Some larger practices have a harder time transforming while smaller practices, which are often rural, can in many cases change more rapidly. As one individual noted, “The uptake in rural areas has been surprisingly high.”

Gaming

One issue mentioned with respect to episodes is the potential for gaming. For example, within the URI episode, there are different procedure and diagnoses codes, (eg, 99214 vs 99213). Also, if an additional diagnosis is added, then fewer URIs will meet the criteria for inclusion as an episode. Patients may be excluded from episode payment due to a comorbid condition. As one provider put it, “Just start adding diagnoses and fewer URIs will make the cut” due to the episode’s exclusion criteria. Arkansas Medicaid’s monitoring of provider use of exclusions has resulted in some minimal changes in

the design of the URI episode, and in more substantial changes to the ADHD episode, including assessments for comorbid conditions. On visit coding, one physician noted that it isn’t a measure of disease severity but rather a measure of what was done during the visit: “It’s a question of whether the patient needs care.” Despite the various incentives under the new system, he added that he hasn’t changed the durations of his office visits (which could affect the appropriate code) “at all.”

Also, this physician pointed out that, if the patient sees him first, then he is the designated PAP. If the patient goes instead to the emergency department first, then he isn’t the PAP even though the patient is still his for purposes of PCMH attribution. In the same way, one provider we interviewed told us off the record of specialists refusing to take referrals in certain cases: those in which they would become PAP and in which it seemed likely that patient behavior would result in the PAP having a negative outcome with respect to episode costs, through no fault of their own. None of the respondents suggested these behaviors were widespread, and ultimately their prevalence is an empirical question.

How Providers Perceive They Can Generate Savings

With respect to savings flowing from changes in practice, there is a much-cited example of antibiotic prescription within the URI episode—often for a common cold—being reduced by 18% to 20%. When the question of excessive prescribing of antibiotics came up in the work group for the URI episode, we were told one doctor “stood up and said not only should they not be prescribed, the patient often shouldn’t even come in [and] now in a PCMH you can get paid for telling them not to come in.”

As another example of changes in practice brought about by the episode component of the model, many physicians who perform a tonsillectomy had been sending tonsils to pathology every time the procedure was performed (one physician notably had a similar practice with respect to placentas in the perinatal episode). These physicians have now changed this behavior, but as a result, 1 interviewee noted, pathologists have “felt the heat, felt the loss.”

Importantly, there was a unified perception across our interviews that providers are practicing similarly for patients regardless of payer and that this is generating spillovers. For example, when asked if providers now practice differently for patients with different payers, the consensus answer was, “absolutely not.”

Payer Reaction

Participating payers we interviewed were generally positive, and viewed the administrative effort and challenges connected with altering provider contracts and payment models as worthwhile. When asked how their relationship with providers has changed as a result of the APII, one insurer stated that the episodic payment component has produced “a different discussion” because of provider concern over the release of performance data related to the initiative.

As far as private payers working closely with the Medicaid program and others in Arkansas state government, one individual we interviewed from the employer community characterized the state’s leadership role as extremely important, noting it was “incredibly serendipitous that the state had the intellectual capital to deliver the right approach” and to use its “big stick”—Medicaid—as the primary driver of reform. When asked if any one insurer or employer could have instituted these reforms on its own without Medicaid’s involvement, the same individual quickly replied, “Never in a million years.”

Though most payers are participating in episodic payment, one challenge with respect to convincing more insurers and self-insured employers to participate relates to the temporal lag in payer incentives between episodes and medical homes. While episodes offer “immediate bang for the buck,” the benefits that flow from the PCMH will likely be “longer-term.” For example, there may be little discernable short-term benefit to employees for paying providers a PMPM for care coordination, even though this may produce better care and value in the future once practice transformation has occurred. As an interviewee explained, it can sometimes be difficult to “think ahead” in this context. Another interviewee from the employer sector noted some employers have expressed reluctance with respect to embracing the PMPM payments, because “They don’t want to be paying for something [practices] should already be doing for nothing.”

An individual from the employer community voiced moderate concern over the level of coordination between PCMH and episodic payment components of the reform, and is “not yet convinced there is strong enough integration.” Another noted anticipation on the part of the business community that reform will lead to greater value: “[Employers] are used to paying an end price for a good quality product and expect to buy healthcare in the same way.”

Questions Going Forward

As phased implementation of the APII continues, many significant questions remain, the most important of which relate to assessing the impact of the program on clinical and economic outcomes. Although anecdotes of savings exist, will episodes actually alter practice, quality outcomes, and costs in a significant way? If there must be ongoing PMPM payments made to incentivize PCMH participation, how suitable is this component of the model as a long-term means of either definitive practice transformation or generating fiscal savings?

Conclusions

The APII could have broad applicability to other states that share Arkansas’ largely rural and fragmented delivery system. The nature of Arkansas’ provider and political landscape may make the APII potentially better suited for domestic export than initiatives in states previously more widely associated with cutting-edge health policy.

The majority of feedback relating to early stakeholder perceptions has been positive. Although our sample of providers was small, representatives of the medical society and hospital association had a broad view of provider reaction across the state. Moreover, while we interviewed relatively few payers, those we did speak with cover the vast majority of Arkansans.

All of those interviewed indicated they are at least receptive to the reforms and many are active participants. If successful, the state’s inclusive, multi-payer approach might provide an important roadmap for how to accelerate payment reform in an era of scarce resources and competing incentives.Author Affiliations: Department of Health Care Policy (MEC, CHM), Healthcare Markets and Regulation Lab (MEC, CHM), Harvard Medical School, Boston, MA; Department of Internal Medicine, University of Michigan Medical School (AMF), Ann Arbor; Arkansas Medicaid (WEG), Little Rock; Arkansas Center for Health Improvement (MWM, JWT), Little Rock; College of Medicine (WEG, JWT), Department of Pediatrics (JWT), College of Public Health, Department of Health Policy and Management (WEG, JWT), University of Arkansas for Medical Sciences (WEG, JWT), Little Rock.

Source of Funding: Robert Wood Johnson Foundation.

Author Disclosures: Drs Golden and Thompson and Mr Motley are affiliated with the University of Arkansas for Medical Sciences (UAMS). Dr Golden is also affiliated with Arkansas Medicaid and the Arkansas Department of Human Services, and has received funding from CMMI. Dr Chernew is the co-editor-in-chief of The American Journal of Managed Care. He has received the Laura and John Arnold Foundation grant: Quantitative Evaluation of Arkansas Payment Improvement Initiative; and the Laura and John Arnold Foundation grant: Supporting Health Reform in Arkansas. Dr Fendrick is a partner of V-BID Health, LLC, co-editor-in-chief of The American Journal of Managed Care, and a member of MEDCAC, as well as a consultant for Amgen, Evergreen Health, Intarcia, Kinetix, Merck, Pfizer, Sanofi, the state of North Carolina, and Zansors. He is also on the advisory board for HCI3 and PCORI, and has completed research for AHRQ, Gary and Mary West Health Policy Center, Laura & John Arnold Foundation, the state of Michigan / CMS, the state of New York, PhRMA, and the RWJ Foundation. Mr Mathis and Mr Motley report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (CHM, JWT, WEG, AMF, MEC); acquisition of data (CHM, JWT); analysis and interpretation of data (CHM, JWT, WEG, AMF); drafting of the manuscript (CHM, JWT, WEG, MM, AMF); critical revision of the manuscript for important intellectual content (CHM, JWT, WEG, MM, AMF, MEC); obtaining funding (MEC); administrative, technical, or logistic support (CHM, JWT, MM); and supervision (JWT).

Address correspondence to: Joseph W. Thompson, MD, MPH, Director, Arkansas Center for Health Improvement, 1401 W Capitol Ave, Ste 300, Victory Building, Little Rock, AR 72201. E-mail: ThompsonJosephW@uams.eduREFERENCES

1. Bitton A, Martin C, Landon BE. A nationwide survey of patient centered medical home demonstration projects. J Gen Intern Med. 2010;25(6):584-592.

2. Rosenthal MB. Beyond pay for performance--emerging models of provider-payment reform. N Engl J Med. 2008;359(12):1197-1200.

3. Smith VK, Gifford K, Ellis E, Rudowitz R, Snyder L. Medicaid in an Era of Health & Delivery System Reform: Results from a 50-State Medicaid Budget Survey for State Fiscal Years 2014 and 2015. Washington, DC: Kaiser Family Foundation; 2014. http://kff.org/medicaid/report/medicaid-in-an-era-of-health-delivery-system-reform-resultsfrom-a-50-state-medicaid-budget-survey-for-state-fiscal-years-2014-and-2015/.

4. Mechanic RE. Opportunities and challenges for episode-based payment. N Engl J Med. 2011;365(9):777-779.

5. O’Byrne TJ, Shah ND, Wood D, et al. Episode-based payment: evaluating the impact on chronic conditions. Medicare Medicaid Res Rev. 2013;3(3). doi:10.5600/mmrr.003.03.a07.

6. Sood N, Huckfeldt PJ, Escarce JJ, Grabowski DC, Newhouse JP. Medicare’s bundled payment pilot for acute and postacute care: analysis and recommendations on where to begin. Health Aff (Millwood) 2011;30(9):1708-1717.

7. Cutler DM, Ghosh K. The potential for cost savings through bundled episode payments. N Engl J Med. 2012;366(12):1075-1077.

8. Silow-Carroll S, Edwards JN, Rodin D, Aligning incentives in medicaid: how Colorado, Minnesota, and Colorado are reforming care delivery and payment to improve health and lower costs. The Commonwealth Fund website. http://www.commonwealthfund.org/publications/case-studies/2013/mar/aligning-incentives-in-medicaid. Published March 1, 2013. Accessed May 2015.

9. de Brantes F, Rastogi A, Painter M. Reducing potentially avoidable complications in patients with chronic diseases: the Prometheus Payment approach. Health Serv Res. 2010;45(6, pt 2):1854-1871.

10. de Brantes F, Rosenthal MB, Painter M. Building a bridge from fragmentation to accountability--the Prometheus Payment model. N Engl J Med. 2009;361(11):1033-1036.

11. Hussey PS, Sorbero ME, Mehrotra A, Liu H, Damberg CL. Episode-based performance measurement and payment: making it a reality. Health Aff (Millwood). 2009;28(5):1406-1417.

12. Arkansas Health System Transformation: State Innovation Plan. Arkansas Health Care Payment Improvement Initiative website. http://www.paymentinitiative.org/referenceMaterials/Documents/SIM%20III.%20%20State%20Innovation%20Plan%202012%2009%2021%20%20FINAL%20-%20TO%20SUBMIT.pdf. Published September 21, 2012. Accessed January 7, 2015.

13. Episodes of Care. Arkansas Health Care Payment Improvement Initiative website. http://www.paymentinitiative.org/episodesOfCare/Pages/default.aspx. Published 2012. Accessed January 7, 2015.

14. Patient Centered Medical Home. Arkansas Health Care Payment Improvement Initiative website. http://www.paymentinitiative.org/referenceMaterials/Documents/AHCPII%20PCMH%20Flyer%204-2013.pdf. Revised April 2013. Accessed January 7, 2015.